- The complexity of marine installations for ocean observing systems has grown significantly in recent years. In a network consisting of tens, hundreds or thousands of marine instruments, manual configuration and integration becomes very challenging. Simplifying the integration process in existing or newly established observing systems would benefit system operators and is important for the broader application of different sensors. This article presents an approach for the automatic configuration and integration of sensors into an interoperable Sensor Web infrastructure. First, the sensor communication model, based on OGC's SensorML standard, is utilized. It serves as a generic driver mechanism since it enables the declarative and detailed description of a sensor's protocol. Finally, we present a data acquisition architecture based on the OGC PUCK protocol that enables storage and retrieval of the SensorML document from the sensor itself, and automatic integration of sensors into an interoperable Sensor Web infrastructure. Our approach adopts Efficient XML Interchange (EXI) as alternative serialization form of XML or JSON. It solves the bandwidth problem of XML and JSON.
In Ocean Observing Systems, the infrastructure for data management, communication and instruments has to be a flexible network, because unanticipated needs may emerge and unprecedented context-aware applications may need to be enabled. For example, a given kind of sensor may be deployed on various platforms such as floats, gliders or moorings, and thus must be integrated with different data acquisition systems. To achieve the interoperability in this infrastructure, the physical instruments must be reliably associated with software components that operate the instruments and manage their data and metadata.
One of the critical issues in an ocean observing system is the communication with the instrument or sensor. In the oceanographic instrumentation sector, there is no standardization of protocols used for the control and configuration of the instruments, and each manufacturer defines the syntax used and the set of commands for each of their instruments. Given the diverse nature of the instruments, it could be difficult to define a universal set of commands. It would, however, be feasible to use a common syntax and a set of common commands in addition to specific commands, depending on the nature of the instrument.
The main objective of standardizing basic installation
processes for new instruments has the main purpose of
reducing operating costs of the observatory. Since each
instrument and manufacturer implements a different
communication protocol and raw data format, it is
necessary to invest much time in the programming of a
software driver that allows the integration of the new
instrument in the observatory network. On the other hand,
there are marine observation platforms such as
oceanographic buoys, where, in addition to the
installation, a configuration tool is needed, and it is
necessary to be able to perform this operation under
extreme conditions. Standardizing these processes
minimizes the risk of failures due to manual
configuration. Another benefit of the standardization
process is the facilitation of maintenance and replacement
of instruments in the observatory and the maintenance of
traceability of the data they generate [
Other benefits are the improved accessibility to data
and interoperability between data sets. However,
interoperability can only be achieved through extensive
use of international standards. They specify regulations
for data access, content, and exchange. The current
situation is characterized by the fact that parallel
approaches have been developed (IEEE 1451, the OGC
set of standards, etc.) but they still lack community
support. This paper introduces the Sensor Web
architecture based on an OGC set of standards as an
example of how interoperable standards help to facilitate
the creation of an infrastructure for sharing
oceanographic observation data and the integration of
sensor data into applications [
II.
OBSERVING SYSTEMS
Oceanographic instruments are traditionally developed by small companies and lack standardization of the protocols for instrument control and configuration, or data retrieval. These instruments are often integrated into an observing system or sensor network, which provides a software infrastructure for functions such as data acquisition, data logging, and data transfer via hardwired or wireless telemetry links.
In order to facilitate the integration of sensors into the
ocean observing system, we propose a Sensor Web
architecture based on the concepts of spatial data
infrastructures and the Sensor Web Enablement (SWE)
framework [
The implementation of the different SWE services, envisioned in the OGC SWE architecture, could depend on the type of observatory platform on which the oceanographic instruments are installed, as shown in Table 1. Therefore, the oceanographic instruments that will be connected to observatories with a continuous and high bandwidth communication channel could implement the full capabilities of the OGC Sensor Observation Service for data access and data push. On the other hand, oceanographic instruments deployed in platforms with a low communication bandwidth or discontinuous link could only implement some of the data push capabilities of the OGC Sensor Observation Service. Moreover, all these instruments will need to implement a configuration service using the OGC Sensor Planning Service, which will allow the observatory operators to configure the oceanographic instruments based on the mission requirements.
To realize the integration of oceanographic instruments into the SWE architecture, the communication between them has to be based on the O&M 2.0 standard that describes a model and an XML Schema to encode data gathered by sensors (archived as well as real-time data). In order to maintain the traceability of the data generated by a specific oceanographic instrument into the SWE architecture, we used the OGC Sensor Model Language (SensorML) to describe the sensor or process that generated that data. Similarly to O&M, the SensorML 2.0 standard describes a model and an XML Schema to encode the metadata of an instrument or a process. Based on SensorML 2.0, we are able to describe the oceanographic instruments, the communication protocol, the observatory platform used for acquisition, and the processes needed by the acquisition system of the observation platform to perform measurements. Such description can be done including an “AggregatedProcess” component into the SensorML document of the instrument as shown in the figure below. A data acquisition system can use such a SensorML instance to translate the communication from the sensor protocol into a standardized OGC SWE communication.
As illustrated in Figure 2, the acquisition system will first use the description given by the “getObservation” process to retrieve data from the instrument. It will then filter the sensor response based on the description found in the “fieldSelector” process. Finally, it will generate the appropriate O&M transaction for the SWE services based on the description of the “insertResult” process.
MANAGEMENT AND INSTRUMENT CONTROL
ARCHITECTURE
Oceanographic instruments that are deployed on cable
observatories, ships or buoys with Ethernet connection,
most of them with limited bandwidth, are implementing a
lightweight SOS [
When the instrument is attached to a 'host' computer
aboard a vehicle, buoy or other observatory platform, the
host uses the protocol description provided inside the
SensorML document to operate the instrument. We use
the OGC PUCK protocol to enable interoperability and
access to the SensorML document of the oceanographic
instruments. OGC PUCK is a simple protocol that makes
it possible for instruments to carry information that
enables sensor networks to use the instrument and its
data. OGC PUCK defines a simple protocol to store and
retrieve information from an instrument over RS-232 and
Ethernet [
Datasheet
Payload
RS 232 or Ethernet
Instrument
One key advantage of implementing OGC PUCK is
that the standard enables the automatic instrument
integration into sensor networks (‘Plug and Work’) in a
very easy manner. Based on the OGC standards, we
implement the interface of oceanographic instruments
with basic standard protocols for sensor detection,
identification, configuration and execution of measuring
operations. We demonstrate the use of this set of
standards on two oceanographic instruments, an optical
and an acoustic instrument shown in Figure 3 [
At connection, the oceanographic instrument is identified by the platform acquisition system using the OGC PUCK protocol. Immediately after the detection, the acquisition system (SWE Bridge) retrieves the SensorML file. Next, the SensorML document is decoded and the SWE Bridge starts to run the set of tasks described in the SensorML, performs measurements, generates the resulting O&M files and inserts them into the standard SOS Server using the observatory specific communication channels as depicted in Figure 5.
We use the Efficient XML Interchange (EXI)
encoding [
This architecture is able to fulfil the central requirements for establishing an interoperable exchange of oceanographic sensor data. While first components are already available, it will be continuously developed and enhanced during this year. Based on currently ongoing evaluation activities of the first available implementations and further emerging requirements, the Sensor Web components will be advanced to a comprehensive suite of tools for sharing oceanographic observation data in an interoperable manner.
This study benefited from the European Union’s Seventh Programme for research, technological development and demonstration, the NeXOS Project No 614102.